Find.pm6

use v6;
=TITLE URL::Find
=SUBTITLE A Perl 6 module to find all the URL's in a set of text.
=head1 DESCRIPTION
=para
By default it will match domain names
that use unicode characters such as http://правительство.рф. To only match ASCII domains use the
:ascii option. It will also find URL's that end in one of the restricted characters, so
`https://www.google.com, ` will pull out `https://www.google.com`. It will find all the URL's in a
text by default, or you can specify a maximum number with the :limit option. By default it will
only find http, https, ftp, git and ssh schemes, but you can specify `:any<1>` to match any schemes
with legal characters..

grammar url {
    has $.ascii = 0;
    has $.any   = 0;
    regex TOP         {  <anyprotocol> '://' <base> [<after>+]? '/'?  }
    token anyprotocol {   <[ a..z A..Z ]> <[ a..z A..Z 0..9 . + - ]>+  }
    token protocol    {:i [http|https|ftp|git|ssh]                     }
    token baseascii   {   [ <[a..z A..Z 0..9 \- . ]> ]+                }
    token base        {   [ <:Number + :Letter + [ . - ]> ]+           }
    token protected   {   <[ $ + ! * ( ) , . ; ? @ = % & # " ' ]>      }
    token allowed     {   \S                                           }
    regex term        {   <allowed>+ <!after <protected>>              }
    token after       {   '/' <term>                                   }
}
class url-actions {
    method TOP ($/) {
        make {
            protocol => $<anyprotocol>.made,
            #url => $<anyprotocol>.made ~ '://' ~ $<base> ~ $<after>.join
        }
    }
    method protocol ($/) {
        make $/.lc;
    }
    method anyprotocol ($/) {
        make $/.lc;
    }
}
#| Accepts a string and returns a list of URL's. Optionally you can specify a limit to the number
#| of URL's returned, or whether you want to only match URL's with ASCII domain names: :ascii<1>
#| Matches only http https ftp git and ssh schemes by default. To match any scheme, use :any<1>
sub find-urls ( Str $string, Num :$limit? is copy, :$ascii?, :$any? ) is export  {
    $limit = ∞ if ! $limit.defined;
    my $base     = $ascii ?? <url:baseascii>   !! <url:base>;
    my $protocol = $any   ?? <url:anyprotocol> !! <url:protocol>;
    my $match = url.parse($string, :actions(url-actions.new)  );
    say $match.made.perl;
    #say $match;
    #my $url-regex = regex { <{ $any ?? &url:anyprotocol !! <protocol }> '://' <$base> [<after>+]? '/'?  };
    #$string.comb($url-regex, $limit);
}
find-urls("https://google.com/something");

=AUTHOR Samantha McVey (samcv) samantham@posteo.net
=LICENSE
This is free software; you can redistribute it and/or modify it under
the Artistic License 2.0.